Wang, Hongli, and Thomas AulignŽ, National Center for
Atmopheric Research
The nonlinearity in the
observation operator is addressed within the framework of an Observation System
Simulation Experiment (OSSE) for cloud-affected satellite radiance
assimilation. This study focuses on various approaches to reduce
observation-minus-first-guess departures and improve the tangent-linear
approximation for the observation operator with the goal of assimilating
cloud-affected radiance observations. The issues addressed are: 1) the ability
of the cloud analysis scheme to retrieve the cloud hydrometeor variables, 2)
the ability of cloud hydrometeor retrievals to reduce the observation
departures and improve the tangent-linear approximation for the observation
operator. The OSSE results show that retrieved cloud condensates produced by
the adjusted GSI cloud analysis package are very similar to the truth in
location and amplitude. Results also show that, compared to the observation
departures with a cloud free background, the observation departures with a
first guess enhanced by the GSI cloud analysis package are significantly
reduced and the tangent-linear approximation for the cloudy radiance
observation operator is improved. The improved first guess using the GSI cloud
analysis package is expected to improve the cloud hydrometeors analysis in a
variational radiance data assimilation system.